Fig. 8From: Hepatic vessel segmentation based on 3D swin-transformer with inductive biased multi-head self-attentionThe different settings to study the effect of kernel size and model bottleneckBack to article page